cecs 694-02 introduction to bioinformatics university of louisville spring 2003 dr. eric rouchka...

CECS 694-02 Introduction to Bioinformatics University of Louisville Spring 2003 Dr. Eric Rouchka

Lecture 2:

Pairwise Sequence Alignment

Eric C. Rouchka, D.Sc.

[email protected]

http://kbrin.kwing.louisville.edu/~rouchka/CECS694/


Sequence Alignment

• Question: Are two sequences related?

• Compare the two sequences, see if they are similar

• Example: pear and tear

• Similar words, different meanings


Biological Sequences

• Similar biological sequences tend to be related

• Information:– Functional– Structural– Evolutionary


Biological sequences

• Common mistake: – sequence similarity is not homology!

• Homologous sequences: derived from a common ancestor


Relation of sequences

• Homologs: similar sequences in 2 different organisms derived from a common ancestor sequence.

• Orthologs: Similar sequences in 2 different organisms that have arisen due to a speciation event. Functionality Retained.



• Paralogs: Similar sequences within a single organism that have arisen due to a gene duplication event.

• Xenologs: similar sequences that have arisen out of horizontal transfer events (symbiosis, viruses, etc)



Image Source: http://www.ncbi.nlm.nih.gov/Education/BLASTinfo/Orthology.html


Edit Distance

• Sequence similarity: function of edit distance between two sequences

P E A R

| | |

T E A R


Hamming Distance

• Minimum number of letters by which two words differ

• Calculated by summing number of mismatches

• Hamming Distance between PEAR and TEAR is 1


Gapped Alignments

• Biological sequences– Different lengths– Regions of insertions and deletions

• Notion of gaps (denoted by ‘-’)

A L I G N M E N T | | | | | | |- L I G A M E N T


Possible Residue Alignments

• Match

• Mismatch (substitution or mutation)

• Insertion/Deletion (INDELS – gaps)


Alignment Scoring Scheme

• Possible scoring scheme:

match: +2

mismatch: -1

indel –2

• Alignment 1: 5 * 2 – 1(1) – 4(2) = 10 – 1 – 8 = 1

• Alignment 2: 6 * 2 – 1(1) – 2 (2) = 12 – 1 – 4 = 7


Alignment Methods

• Visual

• Brute Force

• Dynamic Programming

• Word-Based (k tuple)


Visual Alignments (Dot Plots)

• Matrix– Rows: Characters in one sequence– Columns: Characters in second sequence

• Filling– Loop through each row; if character in row, col

match, fill in the cell– Continue until all cells have been examined


Example Dot Plot


Noise in Dot Plots

• Nucleic Acids (DNA, RNA)– 1 out of 4 bases matches at random

• Stringency– Window size is considered– Percentage of bases matching in the

window is set as threshold


Reduction of Dot Plot Noise

Self alignment of ACCTGAGCTCACCTGAGTTA


Information Inside Dot Plots

• Regions of similarity: diagonals

• Insertions/deletions– Can determine intron/exon structure

• Repeats and Inverted Repeats– Inverted repeats = reverse complement– Used to determine folding of RNA molecules


Insertions/Deletions


Repeats/Inverted Repeats


Comparing Genome Assemblies


Chromosome Y self comparison


Available Dot Plot Programs

• Vector NTI software package (under AlignX)



• Vector NTI software package (under AlignX)

GCG software package:• Compare http://www.hku.hk/bruhk/gcgdoc/compare.html

• DotPlot+ http://www.hku.hk/bruhk/gcgdoc/dotplot.html • http://www.isrec.isb-sib.ch/java/dotlet/Dotlet.html• http://bioweb.pasteur.fr/cgi-bin/seqanal/dottup.pl • Dotter (http://www.cgr.ki.se/cgr/groups/sonnhammer/Dotter.html)



Dotlet (Java Applet) http://www.isrec.isb-sib.

ch/java/dotlet/Dotlet.html


Available Dot Plot ProgramsDotter (http://www.cgr.ki.se/cgr/groups/sonnhammer/Dotter.html)


Available Dot Plot ProgramsEMBOSS DotMatcher, DotPath,DotUp



GCG software package:• Compare http://www.hku.hk/bruhk/gcgdoc

/compare.html

• DotPlot+ http://www.hku.hk/bruhk/gcgdoc/dotplot.html

• DNA strider• PipMaker


Dot Plot References

Gibbs, A. J. & McIntyre, G. A. (1970). The diagram method for comparing sequences. its use with amino acid and nucleotide sequences. Eur. J. Biochem. 16, 1-11.

Staden, R. (1982). An interactive graphics program for comparing and aligning nucleic-acid and amino-acid sequences. Nucl. Acid. Res. 10 (9), 2951-2961.


Determining Optimal Alignment

• Two sequences: X and Y– |X| = m; |Y| = n– Allowing gaps, |X| = |Y| = m+n

• Brute Force

• Dynamic Programming


Brute Force

• Determine all possible subsequences for X and Y– 2m+n subsequences for X, 2m+n for Y!

• Alignment comparisons– 2m+n * 2m+n = 2(2(m+n)) = 4m+n comparisons

• Quickly becomes impractical


Dynamic Programming

• Used in Computer Science

• Solve optimization problems by dividing the problem into independent subproblems


Dynamic Programming

• Sequence alignment has optimal substructure property– Subproblem: alignment of prefixes of two

sequences

– Each subproblem is computed once and stored in a matrix


Dynamic Programming

• Optimal score: built upon optimal alignment computed to that point

• Aligns two sequences beginning at ends, attempting to align all possible pairs of characters


Dynamic Programming

• Scoring scheme for matches, mismatches, gaps

• Highest set of scores defines optimal alignment between sequences

• Match score: DNA – exact match; Amino Acids – mutation probabilities


Dynamic Programming

• Guaranteed to provide optimal alignment given:– Two sequences– Scoring scheme


Steps in Dynamic Programming

Initialization Matrix Fill (scoring) Traceback (alignment)


DP Example

Sequence #1: GAATTCAGTTA; M = 11

Sequence #2: GGATCGA; N = 7

s(aibj) = +5 if ai = bj (match score)

s(aibj) = -3 if aibj (mismatch score)

w = -4 (gap penalty)


View of the DP Matrix

• M+1 rows, N+1 columns


Global Alignment(Needleman-Wunsch)

• Attempts to align all residues of two sequences

• INITIALIZATION: First row and first column set

• Si,0 = w * i

• S0,j = w * j


Initialized Matrix (Needleman-Wunsch)


Matrix Fill(Global Alignment)

Si,j = MAXIMUM[

Si-1, j-1 + s(ai,bj) (match/mismatch in the

diagonal),

Si,j-1 + w (gap in sequence #1),

Si-1,j + w (gap in sequence #2)

]


Matrix Fill (Global Alignment)

• S1,1 = MAX[S0,0 + 5, S1,0 - 4, S0,1 - 4] = MAX[5, -8, -8]



• S1,2 = MAX[S0,1 -3, S1,1 - 4, S0,2 - 4] = MAX[-4 - 3, 5 – 4, -8 – 4] =

MAX[-7, 1, -12] = 1


Filled Matrix (Global Alignment)


Trace Back (Global Alignment)

• maximum global alignment score = 11 (value in the lower right hand cell).

• • Traceback begins in position SM,N; i.e. the

position where both sequences are globally aligned.

• • At each cell, we look to see where we move

next according to the pointers.


Trace Back (Global Alignment)


Checking Alignment Score

G A A T T C A G T T A| | | | | |G G A – T C – G - — A

+ - + - + + - + - - +5 3 5 4 5 5 4 5 4 4 5

5 – 3 + 5 – 4 + 5 + 5 – 4 + 5 – 4 – 4 + 5 = 11


Local Alignment

• Smith-Waterman: obtain highest scoring local match between two sequences

• Requires 2 modifications:– Negative scores for mismatches– When a value in the score matrix becomes

negative, reset it to zero (begin of new alignment)


Local Alignment Initialization

• Values in row 0 and column 0 set to 0.


Matrix Fill(Local Alignment)

Si,j = MAXIMUM[

Si-1, j-1 + s(ai,bj) (match/mismatch in the diagonal),

Si,j-1 + w (gap in sequence #1),

Si-1,j + w (gap in sequence #2),

0

]


Matrix Fill(Local Alignment)

• S1,1 = MAX[S0,0 + 5, S1,0 - 4, S0,1 – 4,0] = MAX[5, -4, -4, 0] = 5


Matrix Fill (Local Alignment)

• S1,2 = MAX[S0,1 -3, S1,1 - 4, S0,2 – 4, 0] = MAX[0 - 3, 5 – 4, 0 – 4,

0] = MAX[-3, 1, -4, 0] = 1


Matrix Fill (Local Alignment)S1,3 = MAX[S0,2 -3, S1,2 - 4, S0,3 – 4, 0] = MAX[0 - 3, 1 – 4, 0 – 4, 0] =

MAX[-3, -3, -4, 0] = 0


Filled Matrix(Local Alignment)


Trace Back (Local Alignment)

• maximum local alignment score for the two sequences is 14

• found by locating the highest values in the score matrix

• 14 is found in two separate cells, indicating multiple alignments producing the maximal alignment score



• Traceback begins in the position with the highest value.

• At each cell, we look to see where we move next according to the pointers

• When a cell is reached where there is not a pointer to a previous cell, we have reached the beginning of the alignment


Maximum Local Alignment

G A A T T C - A

| | | | |

G G A T – C G A

+ - + + - + - +

5 3 5 5 4 5 4 5

G A A T T C - A

| | | | |

G G A – T C G A

+ - + - + + - +

5 3 5 4 5 5 4 5


Scoring Matrices

• match/mismatch score – Not bad for similar sequences– Does not show distantly related sequences

• Likelihood matrix– Scores residues dependent upon likelihood

substitution is found in nature– More applicable for amino acid sequences


Percent Accepted Mutation (PAM or Dayhoff) Matrices

• Studied by Margaret Dayhoff

• Amino acid substitutions – Alignment of common protein sequences– 1572 amino acid substitutions– 71 groups of protein, 85% similar

• “Accepted” mutations – do not negatively affect a protein’s fitness



• Similar sequences organized into phylogenetic trees

• Number of amino acid changes counted

• Relative mutabilities evaluated

• 20 x 20 amino acid substitution matrix calculated



• PAM 1: 1 accepted mutation event per 100 amino acids; PAM 250: 250 mutation events per 100 …

• PAM 1 matrix can be multiplied by itself N times to give transition matrices for sequences that have undergone N mutations

• PAM 250: 20% similar; PAM 120: 40%; PAM 80: 50%; PAM 60: 60%


PAM1 matrixnormalized probabilities multiplied by 10000

Ala Arg Asn Asp Cys Gln Glu Gly His Ile Leu Lys Met Phe Pro Ser Thr Trp Tyr Val A R N D C Q E G H I L K M F P S T W Y V A 9867 2 9 10 3 8 17 21 2 6 4 2 6 2 22 35 32 0 2 18 R 1 9913 1 0 1 10 0 0 10 3 1 19 4 1 4 6 1 8 0 1 N 4 1 9822 36 0 4 6 6 21 3 1 13 0 1 2 20 9 1 4 1 D 6 0 42 9859 0 6 53 6 4 1 0 3 0 0 1 5 3 0 0 1 C 1 1 0 0 9973 0 0 0 1 1 0 0 0 0 1 5 1 0 3 2 Q 3 9 4 5 0 9876 27 1 23 1 3 6 4 0 6 2 2 0 0 1 E 10 0 7 56 0 35 9865 4 2 3 1 4 1 0 3 4 2 0 1 2 G 21 1 12 11 1 3 7 9935 1 0 1 2 1 1 3 21 3 0 0 5 H 1 8 18 3 1 20 1 0 9912 0 1 1 0 2 3 1 1 1 4 1 I 2 2 3 1 2 1 2 0 0 9872 9 2 12 7 0 1 7 0 1 33L 3 1 3 0 0 6 1 1 4 22 9947 2 45 13 3 1 3 4 2 15 K 2 37 25 6 0 12 7 2 2 4 1 9926 20 0 3 8 11 0 1 1M 1 1 0 0 0 2 0 0 0 5 8 4 9874 1 0 1 2 0 0 4 F 1 1 1 0 0 0 0 1 2 8 6 0 4 9946 0 2 1 3 28 0 P 13 5 2 1 1 8 3 2 5 1 2 2 1 1 9926 12 4 0 0 2 S 28 11 34 7 11 4 6 16 2 2 1 7 4 3 17 9840 38 5 2 2 T 22 2 13 4 1 3 2 2 1 11 2 8 6 1 5 32 9871 0 2 9 W 0 2 0 0 0 0 0 0 0 0 0 0 0 1 0 1 0 9976 1 0 Y 1 0 3 0 3 0 1 0 4 1 1 0 0 21 0 1 1 2 9945 1 V 13 2 1 1 3 2 2 3 3 57 11 1 17 1 3 2 10 0 2 9901


Log Odds Matrices

• PAM matrices converted to log-odds matrix– Calculate odds ratio for each substitution

• Taking scores in previous matrix• Divide by frequency of amino acid

– Convert ratio to log10 and multiply by 10– Take average of log odds ratio for converting A to B and

converting B to A– Result: Symmetric matrix– EXAMPLE: Mount pp. 80-81


PAM250 Log odds matrix


Blocks Amino Acid Substitution Matrices (BLOSUM)

• Larger set of sequences considered

• Sequences organized into signature blocks

• Consensus sequence formed– 60% identical: BLOSUM 60– 80% identical: BLOSUM 80


Nucleic Acid Scoring Matrices

• Two mutation models:– Uniform mutation rates (Jukes-Cantor)– Two separate mutation rates (Kimura)

• Transitions• Transversions


DNA Mutations

A

C

G

T

PURINES: A, GPYRIMIDINES C, T

Transitions: AG; CTTransversions: AC, AT, CG, GT


PAM1 DNA odds matrices

A. Model of uniform mutation rates among nucleotides. A G T C

A 0.99 G 0.00333 0.99 T 0.00333 0.00333 0.99 C 0.00333 0.00333 0.00333 0.99

B. Model of 3-fold higher transitions than transversions. A G T C

A 0.99 G 0.006 0.99 T 0.002 0.002 0.99 C 0.002 0.002 0.006 0.99


PAM1 DNA log-odds matrices

A. Model of uniform mutation rates among nucleotides.A G T CA 2 G -6 2 T -6 -6 2 C -6 -6 -6 2

B. Model of 3-fold higher transitions than transversions.A G T C

A 2 G -5 2 T -7 -7 2 C -7 -7 -5 2


Linear vs. Affine Gaps

• Gaps have been modeled as linear

• More likely contiguous block of residues inserted or deleted– 1 gap of length k rather than k gaps of length 1

• Scoring scheme should penalize new gaps more


Affine Gap Penalty

• wx = g + r(x-1)

• wx : total gap penalty; g: gap open penalty; r:

gap extend penalty;x: gap length• • gap penalty chosen relative to score matrix

– Gaps not excluded– Gaps not over included– Typical Values: g=-12; r = -4


Affine Gap Penalty and Dynamic Programming

Mi, j = max {

Di - 1, j - 1 + subst(Ai, Bj)

Mi - 1, j - 1 + subst(Ai, Bj)

Ii - 1, j - 1 + subst(Ai, Bj)

Di, j = max {

Di , j - 1 - extend

Mi , j - 1 - open

Ii, j = max {

Mi-1 , j - open

Ii-1 , j - extend


Drawbacks to DP Approaches

• Compute intensive

• Memory Intensive

• O(n2) space, between O(n2) and O(n3) time


Alternative DP approaches

• Linear space algorithms Myers-Miller

• Bounded Dynamic Programming

• Ewan Birney’s Dynamite Package– Automatic generation of DP code


Significance of Alignment

• Determine probability of alignment occurring at random– Sequence 1: length m– Sequence 2: length n

• Random sequences:– Alignment follows Gumbel Extreme Value

Distribution


Gumbel Extreme Value Distribution

• http://roso.epfl.ch/mbi/papers/discretechoice/node11.html


Probability of Alignment Score

• Expected # of alignments with score at least S (E-value):

E = Kmn e-λS

– m,n: Lengths of sequences– K ,λ: statistical parameters dependent

upon scoring system and background residue frequencies


Converting to Bit Scores

A raw score can be normalized to a bit score using the formula:

• The E-value corresponding to a given bit score can then be calculated as:


P-Value

• P-Value: probability of obtaining a given score at random

P = 1 – e-E

Which is approximately e-E


Significance of Ungapped Alignments

• PAM matrices are 10 * log10x

• Converting to log2x gives bits of information

• Converting to logex gives nats of information


Quick Calculation

• If bit scoring system is used, significance cutoff is:

log2(mn)


Example (p110)

• 2 Sequences, each 250 amino acids long

• Significance: – log2(250 * 250) = 16 bits


Example (p110)

• Using PAM250, the following alignment is found:

• F W L E V E G N S M T A P T G• F W L D V Q G D S M T A P A G


Example (p110)

• Using PAM250 (p82), the score is calculated:

• F W L E V E G N S M T A P T G• F W L D V Q G D S M T A P A G

• S = 9 + 17 + 6 + 3 + 4 + 2 + 5 + 2 + 2 + 6 + 3 + 2 + 6 + 1 + 5 = 73


Significance Example

• S is in 10 * log10x -- convert to a bit score:• • S = 10 log10x

• S/10 = log10x

• S/10 = log10x * (log210/log210)

• S/10 * log210 = log10x / log210

• S/10 * log210 = log2x

• 1/3 S ~ log2x

• • S’ ~ 1/3S


Significance Example

• S’ = 1/3S = 1/3 * 73 = 24.333 bits

• Significance cutoff = 16 bits

• Therefore, this alignment is significant


Estimation of E and P

• For PAM250, K = 0.09; = 0.229

• Using equations 30 and 31:

• S’ = 0.229 * 73 – ln 0.09 * 250 * 250

• S’ = 16.72 – 8.63 = 8.09 bits

• P(S’ >= 8.09) = 1 – e(-e-8.09) = 3.1* 10-4


Significance of Gapped Alignments

• Gapped alignments use same statistics

and K cannot be easily estimated

• Empirical estimations and gap scores determined by looking at random alignments


Bayesian Statistics

• Built upon conditional probabilities

• Used to derive joint probability of multiple events or conditions


Bayesian Statistics

• P(B|A): Probability of condition B given condition A is true

• P(B): Probability of condition B, regardless of condition A

• P(A, B): Joint probability of A and B occurring simultaneously


Bayesian Statistics

• A has two substates: A1, A2

• B has two substates: B1, B2

• P(B1) = 0.3 is known

• P(B2) = 1.0 – 0.3 = 0.7

• These are marginal probabilities


Joint Probabilities

• Bayes Theorem:

– P(A1,B1) = P(B1)P(A1|B1)– P(A1,B1) = P(A1)P(B1|A1)


Bayesian Example

• Given:– P(A1|B1) = 0.8; P(A2|B2) = 0.7

• Then– P(A2|B1) = 1.0 – 0.8 = 0.2; P(A1|B2) = 1.0 – 0.7 = 0.3

• AND– P(A1,B1) = P(B1)P(A1|B1) = 0.3 * 0.8 = 0.24– P(A2,B2) = P(B2)P(A2|B2) = 0.7 * 0.7 = 0.49– …


Posterior Probabilities

• Calculation of joint probabilities results in posterior probabilities

– Not known initially– Calculated using

• Prior probabilities• Initial information


Applications of Bayesian Statistics

• Evolutionary distance between two sequences (Mount, pp 122-124)

• Sequence Alignment (Mount, 124-134)

• Significance of Alignments (Durbin, 36-38)

• Gibbs Sampling (Covered later)


Pairwise Sequence Alignment Programs

• Blast 2 Sequences– NCBI– word based

sequence alignment

• LALIGN– FASTA package– Mult. Local

alignments

• needle– Global

Needleman/Wunsch alignment

• water– Local

Smith/Waterman alignment


Various Sequence Alignments

Wise2 -- Genomic to protein

Sim4 -- Aligns expressed DNA to genomic sequence

spidey -- aligns mRNAs to genomic sequence

est2genome -- aligns ESTs to genomic sequence

cecs 694-02 introduction to bioinformatics university of louisville spring 2003 dr. eric rouchka...

Documents

similar sequences

c t slide

eric rouchka alignments

t c g g

eric rouchka lecture

sequence alignment eric

homologous sequences

e n t slide